When building a multiple linear regression model, if we wish to predict the profit margin based on the list price and color of a car, we would $q,$

create a single dummy variable for the color and use the dummy variable in our model.

be better off using a classification decision tree instead.

not be able to create a regression model since the color is categorical data.

have to create dummy binary variables for each car color and use those in the model.